Professional Documents
Culture Documents
Reshaping The Future With NFV and SDN - ADL - May15
Reshaping The Future With NFV and SDN - ADL - May15
May 2015
Content
1. The challenge and the vision
12
16
19
21
8. Conclusion
23
Preface
Network innovation, in the form of Network Functions Virtualization (NFV) and Software Defined
Networking (SDN), may seem to be history repeating itself. The cycle is familiar. A significant new
technology arrives on the scene and threatens to shake up the status quo. Sometimes the
technology is led by customer demand. At other times, it results from a supply-side push. From the
10,000-foot perspective, NFV and SDN may not look much different from any other supply-side
technology push. However, the scope and scale of transformational change that lies ahead cannot
be overstated. In our view, these technologies could be as significant as the introduction of IP
networks themselves.
The late 90s saw mass deployment of IP networking. At the time, large corporations were tired of
expensive switched-networking technologies and needed lower-cost, flexible networking solutions.
On the residential front, access to free Internet content became the new killer application. Driven
by customer demand, carriers overlaid IP networks on to legacy technologies such as ATM1. This
customer-driven network expansion has continued for nearly two decades, first overlaid and then
alongside legacy communications systems.
As a result, IP modernization in carrier networks has been restrained by the need to accommodate
legacy network functions and management systems. This has meant, the system gains possible
from full IP transformation and modernized operations have eluded most carriers. NFV and SDN,
however, could provide both the impetus and the opportunity to change. Carriers are now seeking
broad network automation and programmability transformation, by taking advantage of the same
technologies and platform designs that have revolutionized the management of applications, data
centers for over-the-top (OTT) and web-scale companies and cloud service providers.2
The dynamics of NFV and SDN adoption are quite different from those that drove mass adoption of
IP networking. The IP networking trend was initiated by a customer pull by enterprises wanting
more efficient IP-optimized services; for lower cost Internet access; more mainstream WAN
networking technology (similar to LAN), and the prospect of dynamic VPN creation and routing
leveraging IP/MPLS technology. However, there are two forces in this new technology cycle. On
one side, enterprises are adopting cloud architectures and IT technologies but these are working
over less flexible carrier IP/VPN connectivity services, so they are increasing their use of over-thetop Internet VPNs. On the other side, carriers are beginning to leverage cloud and IT technologies
to build the type of network service flexibility, dynamics and operational efficiency required to serve
their customers needs in the cloud age. So there is a push and a pull aspect to this new era,
which can be attributed to IP networking extension on one side and IT system extension on the
other to form a common IP-IT platform.
1 ATM refers to Asynchronous Transfer Mode technologies used for reliable circuit transport of communications signals.
2 Refers to companies that provide services delivered over public IP networks such as YouTube, Netflix, Whatspp, Viber and Line, as well as those providing cloud
infrastructure to others, such as AWS, Azure, Google Compute, Rackspace, Softlayer and so forth.
3
While the industry has yet to align behind a single vision of future networks, most see NFV and
SDN as central to the future. Therefore, industry R&D efforts are being driven by the promise of
lower operating expenses, along new value streams from new service portfolios for enterprises,
wholesale customers, and emerging web customers served from highly automated operations.
Every previous major technology cycle in the telecom industry has created winners and losers with
most ultimately driving higher demand and expanding revenues across the whole industry. Will this
hold true for NFV and SDN? In our view, there are many opportunities for the expansion of value in
the telecom sector, especially for carriers that purposefully embrace this transformation. But
carriers that fail to internalize these changes are likely to find themselves left behind with
uncompetitive and potentially untenable business and operating models.
In our view, there are many opportunities for the expansion of value in the telecom sector,
especially for carriers that purposefully embrace this transformation. But carriers that fail to
internalize these changes are likely to find themselves left behind with uncompetitive and
potentially untenable business and operating models.
This joint paper examines the economic impact of NFV and SDN on carriers. It aims at a rational
and informed analysis reflecting neither hype curve peak nor trough views and highlights
the strategic and business implications of the change. While the focus of the analysis is Europe,
many of the conclusions may also apply to carriers operating in competitive markets in North
America and Asia Pacific. The report is offered to a technical audience, business managers and
clued-in financial analysts and investors to facilitate debate about new strategic development paths
enabled by these technologies.
Arthur D. Little and Bell Labs have collaborated on this report because we recognize that bringing
NFV and SDN into carrier networks while vital is a major undertaking. Therefore, before
investment priorities are set, a rational big picture economic and strategic assessment is required
to determine the most appropriate long-term goals. Both organizations have a long heritage in
serving clients through application of technology and analysis, and together we bring a unique
perspective on the future of carrier networks.
We hope it will be an eye-opening read.
Sincerely
Ignacio GarciaAlves Marcus Weldon
Global Chairman and CEO
President of Bell Labs and CTO
Arthur D. Little
Alcatel Lucent
4
Abstract
A geographically diverse set of Tier 1 carriers is driving initiatives to standardize critical components
of network functions virtualization (NFV) and software-defined networks (SDN). Led by AT&T,
Deutsche Telekom, NTT, Telefonica and Verizon, among others, these carriers hope to bring new
technologies first developed and proven in the data center into their networks. The excitement over
NFV and SDN has created renewed interest in the networking business and has fanned debate
regarding potential winners and losers. So far, however, there has been only limited analysis of the
potential economic and competitive impact of these technologies on the network, operations and
the bottom-line.
This joint white paper by Arthur D. Little and Bell Labs aims to fill that gap. It offers our views on the
network of the future and discusses the value of fusing carrier networks with the cloud ecosystem,
as enabled by these technologies. These changes are both an opportunity and a necessity for
telecom operators and for those that get it right, it will be a game changer.
5
Pre
2007
Apr
2008
Oct
2009
Concept of SDN
emerges from
research on
active and
programmable
networks
First open
source code of
OpenFlow
protocol for
Campus
networking
announced
First open
source code
for vSwitch
announced
2006
2007
2008
2009
Nov
2013
networks by
VMWare
Jan
2012
DTAG Basic
European
Network
2010
2011
Apr
2012
Feb
2014
First release
OpenDaylight
code
Hydrogen
Google
publicises
use of
OpenFlow
2012
2013
June Formal
Discussions
2012 on NFV
Significant venture
capital funding begins
for companies SDN
and network
virtualization (e.g.,
Nicira Networks)
July
2007
Mar
2011
Oct
2012
Jul Acquisition
2012 of Nicira
April
2013
2014
The Linux
Foundation
found the
project
OpenDaylight,
an open source
framework to
accelerate SDN
NFV adoption
ETSI-ONF
collaboration
formalized on
SDN support
of NFV
Mar
2014
A
s much as 14 billion euros in direct network operating
expense (OPEX) savings
A
further 25 billion euros of non-network savings from much
needed changes in the operating model of the industry
leading to a total savings of 39 billion euros, or 26 percent of
overall OPEX
Capturing the potential of network transformation requires not
only technical skill, but leadership in making significant changes
to current business and operating practices. The ICT industry
as a whole telecom, IT and device manufacturers as well
as governments have all played important roles in creating the
conditions for the Internet, cloud and mobile device revolution.
Now its vital for the telecommunications industry to advance
networking to a whole new level of programmability to facilitate
the continued development of the web in the cloud era and
to ensure the industrys relevance in the cloud ecosystem.
Domain 2.0
The market is moving rapidly towards the adoption cycle along two dimensions; the customer pitch and purchasing
Source: Public domain documentation
3 Programmable in the sense that the network can be customized and controlled through flexible APIs; automated in the sense that the network can fulfill these requests
without manual operations.
Network automation
The other side of the NFV and SDN coin is network automation.
Realizing systemic automation requires systemic process
simplification and redefinition of the service delivery model.
These advances enable dynamic pricing, policy-based ondemand resource access, service instantiation, assurance,
analytics, and charging at cloud speed.
In this new model, the existing network is fully transitioned to
a streamlined, all-IP/Ethernet/optical network architecture from
the access to the core. Moreover, wherever its practical and
cost effective, service-delivery platforms and network functions
will be decoupled from special-purpose hardware platforms and
implemented as virtualized software, deployed on commercial,
general-purpose computing hardware, in much the same way as
typical cloud services. These profound changes will also facilitate
the drive toward highly converged platforms and integrated
control and operations functions across fixed and mobile
operations.6 Supporting this new architecture will require a new
approach to OSS and BSS.7
4 Connectivity services are typically private network and virtual private network services of various types for enterprise and data center interconnection, as well as
enterprise and consumer Internet access and communications services across fixed and mobile networks.
5 Refers to being able to expand programmable services across partner operators, with the ability to offer the same service capabilities independently of whichever
partner owns the serving network.
6 For example, converged IP edge functions (such as IP edge routers, security, CDN, video optimizations and value-added services) consolidated at edge cloud nodes,
and IP core functions (IMS/service platforms, policy, subs data, BSS/OSS, etc.) in more centralized data centers.
7 OSS refers to operations support systems, and BSS stands for business support systems computer systems used by service providers to manage their networks.
They support management functions such as network inventory, service provisioning, network configuration and fault management, among others.
10
11
Figure 3: Future architecture of the carrier network enabled by SDN and NFV
vApp
vApp
vApp
vNF
Enterprise/ vApp
Private Cloud
vApp
vApp
Public
Cloud/CDNs
vNF
Streamlined
IP/Optical network
Optimized wireline/
Wireless access
vApp
vApp
vUC
vM2M
vHSS
vDNS
NaaS
vAAA
PaaS
vIMS
IaaS
vPCC
12
vUCaaS
vNF
vSecurity
SDN control
NFV management and orchestration
Converged applications
Telco
Central Cloud
Services/vNFs
Dynamic
software-defined
VPNs
Network Core
All-IP/Optical
vApp
vApp
vEPC
vBBU
vCDN
vIDS/IPS
vDPI
NaaS
vBNG
Telco
Distributed vCPE
Services/vNFs
PaaS
Network Access/Edge
All IP/Ethernet/
Optical
E2E QoS
vNF
Network APIs
IaaS
NFaaS
Vendors
Virtual Functionality
Applications
Extended Network
Services
End user
equipment
Network Core
Network Edge
Functions
With these enhanced networks, along with increased selfprovisioning, enterprises gain greater control over the real-time
bandwidth use of their applications. They can also use this
flexibility to offer a better user experience to both internal and
external customers by tailoring the quality of service to the
needs of individual applications and by dynamically scaling
service capacity when needed. In each case, elasticity and
bandwidth control will create monetization opportunities.
Software-defined VPNs (SD-VPNs) the virtualized alternative
(or complement) to traditional VPNs for enterprises
incorporate many network virtualization and programmability
capabilities. They enable rapid site turn-up and secure
connectivity over any IP access, along with agile, policy-driven
automated provisioning, efficient security management, and
enterprise self-administration. They also include programmability
that enhances an operators ability to offer service trials. Such
try-before-you-buy offers promote greater market acceptance
through broader service awareness, and they can accelerate
revenue capture through faster service adoption.
User control of available network capacity, through elastic
scaling and bandwidth management, is the essence of policyand application-driven networking which is used to create
and redefine customized, context-dependent services in real
time. The utility of this elastic capability can be bolstered through
network analytics that dynamically monitor network state
and quickly adapt service delivery accordingly. These dynamic
services can be monetized using innovative pricing models
that reflect the enhanced user experience and are optimized
with respect to current network utilization levels and demand
elasticity.
For a typical Tier 1 operator, we estimate that these dynamic
services, together with the pull-through services described in
the following section, could generate an increase of 10 percent
to 15 percent in existing enterprise customer data revenues.
A significant new market segment for dynamically enabled
services can be found among web-scale players and large
global enterprises. Thats because SDN-enabled networking
provides the capability to rapidly deploy new multi-national sites
(even temporary sites), adapting their inter-connectivity and
management to dynamically changing workloads through policybased control.
13
14
New collaborations
A rich new set of business models, based on flexible forms of
network disaggregation, will be enabled, through the mix of
on-demand access to connectivity, along with exposing core
network functionalities to other entities. This change is a far cry
from open-access provisions mandated by market liberalization
regulations. Rather, it is a market-driven approach that will
enable a much bigger network services ecosystem. Carriers
and third parties will work together actively to devise and enrich
service offerings, leveraging network-function-as-a-service
(NFaaS) capabilities from carriers, as well as federated access
and network services accessed with open network APIs.
The implications are both promising and challenging for carriers.
Carriers will be able to buy into each others dynamic VPNs
and NFaaS offerings to expand their services beyond their
traditional geographic footprints. The ability to link these virtual
network capabilities on demand will spur a wave of new product
introductions to take advantage of new operating synergies and
new business opportunities among carriers, network-equipment
vendors and cloud-service providers. This activity will be much
like current trends among OTT and web-scale companies and
their partner ecosystems. When done right, these collaborations
can create promising new value for customers and partners.
But in the process, they may accelerate erosion of traditional
15
23
6
17
59
11
8
5
5
111
44
91
67
Technology onboarding
Technology onboarding, as defined here, refers to broadly
employing automated NFV and SDN capabilities to transform
the carrier network. This process is achieved by adopting
software-based network functions, wherever practical and
cost effective. These functions are deployed on commercial
off-the-shelf hardware, linked by dynamically configured
Combined
OpEX 20131
NW OPEX
Combined
impact
8 Across the 35 European countries (including Turkey) covered in this study, revenue from services reached 275 billion euros (adjusted 250 billion excluding ICX and
handset sales) in 2013. Operating costs represented 60 percent of overall adjusted revenues, equal to 170 billion euros (adjusted 150 billion excluding ICX and handset
sales). Across Europe, the industry employed a staff of 665,000, excluding external contractors.
9 Management of the network infrastructure, including all network-related operations and IT, is the single largest OPEX category and is very labor intensive. In 2013, the
cost base was 50 billion euros or 33 percent of adjusted costs and employed a staff of 180,000, equal to 27 percent of headcount. These costs increase further by 8
billion, or 6 percent of adjusted costs and 84,000 staff (13 percent), if the costs of installing fixed-line customer premises equipment is also considered in the scope of
the network. Fixed network operators outspend their mobile counterparts by approximately 37 billion euros. Moreover, they employ six times as many staff. Of a total
of 460,000 employees, network staff is over 50 percent, with approximately 230,000 people. In contrast, all mobile operators employed 205,000 staff, of which only
35,000 performed network-related functions.
16
Euro billions
-25 bn
-14 bn
58,7
91
7,4
2,7
9
2
6
1,7
3,0
3,2
1,2
1,3
NW OpEX Access/
2013
Agregation
5
3
0,7
Service
platforms
6
3
3
3
2
1,8
0,2
1,0 0,5
1,2
0,4
0,6 0,3
2
2 0
67
0,8
0,5 0,2 44,4
0,1
Core
Labour
G&A
IT
Facilities
Final Adj.
OpEx
10 Across the 35 European reference countries, 360 million fixed lines and 775 million mobile SIMs were in service in 2013.
17
Euro billions
Euro billions
-25%
-27%
76
74
8
Technology
on-boarding
Adj OPEX
2013
NW savings
66
Adj OPEX
70
13
12
54
Rewrite
operating
model
Final Adj
OPEX
Adj OPEX
2013
Technology
on-boarding
NW savings
Adj OPEX
Rewrite
operating
model
57
Final Adj
OPEX
enable savings of 23 euros per fixed line and only 8 euros per mobile SIM, whereas rewriting the operating model will generate a
further 32 euros per fixed line and 17 euros per mobile SIM (see Figures 7 and 8).
The following section discusses how the industry and individual operators should respond to the opportunity.
205
23
182
32
150
-25%
98
Adj OPEX
2013
Technology
on-boarding
NW savings
Adj OPEX
Rewrite
operating
model
18
Final Adj
OPEX
Adj OPEX
2013
Technology
on-boarding
NW savings
90
Adj OPEX
17
74
Rewrite
operating
model
Final Adj
OPEX
11 The telecom industry, threatened by OTT and web-based service providers, has argued that they must absorb the massive costs of developing and operating the global
networks, while OTT players benefit from these without apparent regard for the traffic they generate.
19
12 This characterization does not preclude that a telco may create an internally focused cloud carrier to centralize management of international network of OpCo.
13 Such an archetype could continue to provide non-access services through white-label cloud carriers or traditional partners XaaS offerings.
20
21
22
8. Conclusion
Network innovation based on NFV and SDN holds tremendous promise for the future of networks. But for this promise to be
realized, carriers must transform their heterogeneous networks into a streamlined, automated whole. To achieve this goal, they
need to seek inspiration from cloud peers, while maintaining the reliability and service quality customers expect. None of that is
simple. However, the prize for carriers that get it right is significant. It includes systemic streamlining of operations, exposure of
network services to high-volume, dynamic consumption by cloud applications and cloud-based enterprises and perhaps even the
opportunity for decommoditization of the network.
Carriers must plan and phase such an extensive network transformation carefully. The process includes iterating frequently and
working closely with both suppliers and lead customers to maximize the benefit and the learning at each phase. It also involves
building up increasingly value-rich solutions that advance toward the ultimate goal. Some key actions include:
A
dapting business models to actively shape marketplace adoption of cloud/IP networking services from enterprise customers as
well as wholesale segments (OTT and web-scale, carriers carriers, IT players, etc.)
E
ngaging openly with other carriers regarding standards alignment to facilitate cloud/IP networking enablement, including
tackling regulatory issues of cross-border data storage and privacy concerns
I nvolving suppliers in the early stages of planning network transformation, and monitoring the emerging cloud networking
ecosystem to identify disruptions in the supplier landscape as this space matures
Developing alliances and joint ventures (similar to airlines) to support the global dynamic cloud network needs of multi-national
corporations and OTT and web-scale companies, based on a shared vision of an automated network services catalogue
evolution
F
ormulating a strategy and a timeline for aggressively retiring legacy services that will facilitate a more rapid transformation to
the cloud/IP network
The opportunities and decisions that carriers now face, given the expected breadth of changes from SDN, NFV, and the cloudnetworking market, make this one of the most exciting, pivotal times in the history of communications networks. A lot is at stake.
And as in any major disruption, there will be winners and losers. We believe that the winners in this case will be the carriers that:
Move early and purposefully
Co-create the future with lead customers and partners
Take the opportunity to not only virtualize and automate their networks, but also thoroughly converge and streamline them
These will be the agile service providers equipped to thrive in the cloud ecosystem.
23
Authors
Authors:
Jess Portal
Salman Ali
Arthur D. Little
portal.jesus@adlittle.com
Arthur D. Little
ali.salman@adlittle.com
Enrique Moral
Kevin A. Sparks
Arthur D. Little
moral.enrique@adlittle.com
Bell Labs
kevin.sparks@alcatel-lucent.com
Enrique Hernandez-Valencia
Subramanian Prakash
Narayan Raman
Bell Labs
narayan.raman@bell-labs.com
Acknowledgement for their support and valuable input: Volker Pfirsching and Anders Johansson
Contacts
If you would like more information or to arrange an informal discussion on the issues raised here and
how they affect your business, please contact:
Austria
Karim Taga
taga.karim@adlittle.com
Italy
Giancarlo Agresti
agrestic.giancarlo@adlittle.com
The Netherlands
Martijn Eikelenboom
eikelenboom.martijn@adlittle.com
Belgium
Gregory Pankert
pankert.gregory@adlittle.com
Japan
Shinichi Akayama
akayama.shinichi@adlittle.com
Nordic
Martin Glaumann
glaumann.martin@adlittle.com
China
Antoine Doyon
doyon.antoine@adlittle.com
Korea
Hoonjin Hwang
hwang.hoonjin@adlittle.com
Singapore
Yuma Ito
ito.yuma@adlittle.com
Czech Republic
Dean Brabec
brabec.dean@adlittle.com
Latin America
Guillem Casahuga
casahuga.guillem@adlittle.com
Spain
Jesus Portal
portal.jesus@adlittle.com
France
Didier Levy
levy.didier@adlittle.com
Switzerland
Clemens Schwaiger
schwaiger.clemens@adlittle.com
Germany
Michael Opitz
opitz.michael@adlittle.com
Malaysia
Thomas Kuruvilla
kuruvilla.thomas@adlittle.com
UK
Richard Swinford
swinford.richard@adlittle.com
India
Srini Srinivasan
srinivasan.srini@adlittle.com
Middle East
Thomas Kuruvilla
kuruvilla.thomas@adlittle.com
USA
Guillem Casahuga
casahuga.guillem@adlittle.com
26
Arthur D. Little
Arthur D. Little has been at the forefront of innovation since
1886. We are an acknowledged thought leader in linking
strategy, innovation and transformation in technology-intensive
and converging industries. We navigate our clients through
changing business ecosystems to uncover new growth
opportunities. We enable our clients to build innovation
capabilities and transform their organization.
Our consultants have strong practical industry experience
combined with excellent knowledge of key trends and
dynamics. Arthur D. Little is present in the most important
business centers around the world. We are proud to serve most
of the Fortune 1000 companies, in addition to other leading
firms and public sector organizations.
For further information please visit www.adl.com
Copyright Arthur D. Little 2015. All rights reserved.
www.adl.com/NFVandSDN